Research On Target Detection Algorithm Of Airborne Forward-looking Scanning Radar

Airborne forward-looking scanning radar can obtain the real beam image of the ground/sea scene in front of the platform and has potential application value in the fields of ground reconnaissance,precise guidance,sea surface monitoring,and regional rescue.However,the azimuth resolution is usually poor due to real beam images being limited by the beam width and range,and it's difficult to extract multiple targets of interest.In particular,it is impossible to obtain a more complete target contour and a lower false alarm rate at the same time.This thesis focuses on the target detection of airborne forward-looking scanning radar,especially the detection of multiple targets at different scales,and studies the detection method based on the image data structure.The main contents are as follows:1.The echo model of the airborne forward-looking scanning radar is established,and the azimuth convolution model is derived through mathematical operations.The illconditioned problem of deconvolution is analyzed,and the super-resolution based on statistical optimization is studied and simulated in this thesis.It provides a basis for subsequent muti-target detection.2.The target detection algorithm based on low rank and sparse decomposition is studied in this thesis.The low rank and sparse decomposition model is established by taking advantage of the low rank of the background and the sparsity of targets in the super-resolution image.The multi-object detection of interest is realized by constructing and solving the objective function.Aiming at the problem that the detection method is sensitive to regularization parameters,and the detection results have a high false alarm rate.A target detection scheme combined with the constant false alarm algorithm is proposed in this thesis,and compared with the traditional target detection algorithm.3.The target detection algorithm based on tensor decomposition is studied in this thesis.Aiming at the problem that the detection algorithm can't balance the accuracy and efficiency,a tensor representation is constructed by image segmentation.And a tensor decomposition function model is established,which transforms the target detection problem of airborne forward-looking scanning radar into an optimization problem of the objective function.Then,the weight factor is designed to suppress the background clutter and the objective function is solved by the alternate direction multiplier method.The multi-target detection of forward-looking scanning radar based on tensor decomposition is realized,and the comparison with the low rank and sparse decomposition method is carried out.The methods proposed in this thesis have been verified by MATLAB simulation.The simulation results show that the methods proposed in this thesis can extract multitarget of interest at different scales in the forward-looking area of the airborne scanning radar effectively.